JPH06152970A - Picture compressing method and picture processor - Google Patents

Abstract

PURPOSE:To attain natural picture restoration by providing a deleting means to delete some of picture elements in an area obtained by an extracting means. CONSTITUTION:This picture compressing method and this picture processor are the method and the device for compressing a picture capable of reproducing delicate change in the picture by a step (101) to read in the picture as plural picture elements, the step (103) to detect the degree of color change between the neighboring picture elements, the step (104) to extract the area in which the color change of the picture element is not steep compared with another area, and the step (107) to delete some of the picture elements in the area obtained by this extracting step.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to an image compression method and an image processing apparatus using the image compression method.

[0002]

2. Description of the Related Art Generally, an image processing system is an image input device for inputting an image of an object, an information processing device for recognizing, analyzing, and processing the input image, and an image output for outputting the processed image. It is equipped with a device.

In such an apparatus, the image is compressed in order to process the sampled image efficiently. The compression means making the image data smaller than the original image data, but in this case, the condition is that the original image can be reproduced with this small amount of image data.

As such a compression method, for an image captured by an image input device, the pixels constituting the image are thinned out according to a certain rule, for example, every other pixel or every two pixels. is there. In this case, at the time of image restoration, a method is adopted in which pixels adjacent to the thinned pixels are copied to the thinned portion to restore the original number of pixels.

[0005]

However, according to the method described above, even in the portion where the color change of the image is large and the portion where the color change is not large are thinned out evenly, the portion where the color change is large when the image is restored. The degree of restoration of is much worse than the degree of restoration of the part where the color change is not significant.

The present invention has been made in view of the above points, and it is an object of the present invention to provide an image compression method having a high degree of restoration and an image processing apparatus employing this image compression method.

[0007]

The present invention adopts the following means in order to solve the above problems. As shown in FIG. 1, the system configuration of the image compression apparatus of the present invention includes (1) input means for reading an image as a plurality of pixels (2) color change detection means (3) for detecting the degree of color change between adjacent pixels. ) Extraction means for extracting an area in which the color change of the pixel is not sharp compared to other areas (4) Deletion means for deleting some of the pixels in the area obtained by the extraction means (5) Storage unit for storing the pixels remaining after deletion (6) A restoration unit for reading and restoring the pixel information stored in the storage unit.

The image compression method of the present invention implemented by such an apparatus is (1) a step of reading an image as a plurality of pixels (2) a color change detection step of detecting the degree of color change between adjacent pixels ( 3) Extraction step for extracting an area in which the pixel color change is not sharp compared to other areas (4) There is a deletion step for deleting some of the pixels in the area obtained in the extraction step.

Here, the image input means can be exemplified by a color image scanner, a color TV camera, and other color image pickup devices. The color change detecting means is configured to detect the R value of the pixel, G
R, G, and B values detected in adjacent pixels are compared with each other by comparing the RGB value detection unit that detects at least one of the R value, the B value, and the RGB value detection unit. It can be designed so as to have a color change degree output section that outputs as a change degree.

The RGB value detecting section detects the value of R ² + G ² + B ² of the pixel, the value of R ² + G ² of the pixel, and the value of R ² + B ² of the pixel as the degree of color change. Detecting, detecting the value of G ² + B ² of a pixel, and other processing such as R value only, G value only, B value only, or n-th power.

In the above-mentioned case, the color change degree output section outputs R ² + G ² + B ^{2 of} the first pixel and R ^{2 of} the second pixel adjacent thereto.
² + G ² + B ² is compared and the difference is output as the degree of color change.

In the above case, the color change degree output unit is configured such that R ² + G ² of the first pixel and R ² + G of the second pixel adjacent to the first pixel R ² + G ² .
² is compared and the difference is output as the degree of color change. In the above case, the color change degree output unit outputs R ² + B ² of the first pixel.
And R ² + B ² of the adjacent second pixel are compared with each other and the difference is output as the color change degree.

[0013] When the color change degree output unit, G ² + B of the second pixel adjacent the G ² + B ² of the first pixel and that
² is compared and the difference is output as the degree of color change. In addition, as another example in the above, for example, the first pixel: R ₁ , G ₁ , B ₁ The second pixel: R ₂ , G ₂ , B ₂ The third pixel: R ₃ , G ₃ , B ₃ Then, {(R ₁ + R ₃ ) / 2} -R ₂ = R _d {(G ₁ + G ₃ ) / 2} -G ₂ = G _d {(B ₁ + B ₃ ) / 2} -B ₂ = from formula of B _d, and an intermediate point of the color change that leads to the third pixel from the first pixel, the difference value R _d of the second pixel, G _d, B _d may calculate the color change degree output unit.

The point is that the color change detection means is not limited to the above, because the color change degree can be represented by some parameter. For example, the color of each pixel may be regarded as lightness, and may be classified into a plurality of, for example, 256 gradations for comparison. When the image is captured by the brightness as described above, an image input unit capable of inputting the brightness of a plurality of gradations, such as a monochrome TV camera, can be used. The color change in this case depends on whether the degree of change of the black and white gradation is large or small. That is, the color change in the present invention is a broad concept including a change in brightness and a change in productivity.

The extracting means has a color change degree comparing section for comparing the color change degree obtained by the color change detecting means with an arbitrarily set reference value. In the color change degree comparing section, the color change degree is calculated. When it is determined that the degree is smaller than the reference value, it can be designed so as to be specified as the color change non-steep area.

Here, the reference value can be set to an arbitrary value in advance by a program, or can be set to an arbitrary value by post-setting from an input means such as a keyboard. When the reference value is increased, the area where the color change is abrupt is reduced, and when the reference value is decreased, the area where the color change is abrupt is increased.

The specification of the steep color change region means that the non-steep color change region in which the color change is not sharp is specified. When _obtaining the values of R _d , G _d , and B _d exemplified above, if the difference value is smaller than the preset reference value, the data of the second pixel is discarded. When the difference value is larger than the preset reference value, the data of the second pixel is left as it is.

Here, the values of R _d , G _d and B _d are respectively compared with a reference value, and if at least one of them is within the reference value (range of error), at least two of them are the reference values. Within, or at least three of which are within the reference value, or R _d + G _d + B
It is determined whether _d = K is within the reference value.

After specifying the steep color change region and the non-steep color change region, for convenience of post-processing, the ID assigning means
An identifier may be assigned to both or one of the identified steep color change region and the steep color change region.

When the non-steep color change region is specified, the deleting means partially deletes pixels in the non-steep color change region. The deletion method is a deletion method according to a certain rule such as every other pixel or every two pixels.

In the sharp color change region, the pixels are not deleted as they are. Image compression is completed by the above deletion.
The compressed image information is stored in the storage means. The restoration means reads out the image information from the storage means and reproduces it.

An identifier is useful for reproduction.
That is, if the read information is an abrupt color change region by the identifier, it may be restored as it is. In the case of the information of the non-steep color change area, the image is restored by copying the pixel and supplementing the deleted portion.

[0023]

According to the present invention, an image is read as a plurality of pixels, the degree of color change between adjacent pixels is detected, a region in which the pixel color change is not sharp compared to other regions is extracted, and this color change is detected. Are thinned out in some pixels in a region where the image is not sharp, and pixels are not thinned out in the region where the color change is sharp.

Therefore, when restoring the image, there is no influence on the restoration of the portion where the color change is abrupt.
The image will be restored more naturally. The image processing apparatus of the present invention is
It is suitable for not only compression / playback of still images but also compression / playback of moving images.

[0025]

【Example】

<Embodiment 1> An embodiment of the present invention will be described below with reference to the drawings.

FIG. 2 shows an example of the image processing apparatus of this embodiment. In FIG. 2, reference numeral 11 is a color image scanner as an input unit 11 that reads an image as a plurality of pixels. A color image analyzer 15 forming a color change detection unit 13 is connected to the input unit 11. The color image analyzer 15 converts the color image signal read from the color image scanner into RGB 3 signals.
It is decomposed into primary color signals, and each density value of the RGB signal is expressed in 256 steps for each pixel.

The color image analyzer 15 is connected to a color change degree output section 17 which constitutes a color change detection section. The color change degree output unit is composed of a calculation unit of the information processing device. The color change degree output unit 17 detects the degree of color change between adjacent pixels among the pixels forming the image in the image captured by the color image analyzer 15. The color change degree output unit 17 compares the RGB values analyzed by the color image analyzer 15 between adjacent pixels and outputs the difference between the two as a color change degree.

Here, in comparison, R ² of a certain pixel is compared.
The value of + G ² + B ² is calculated, and at the same time, the value of R ′ ² + G ′ ² + B ′ ² of the pixel adjacent to the pixel is calculated. Then, the difference K between the two is calculated by the following equation.

K = (R ² + G ² + B ² ) − (R ′ ² + G ′ ² +
B ^'2) This K is the color change degree in the present invention. Subsequent to the color change degree output unit 17, a region extraction unit 19 is provided that extracts a region where the color change of pixels is not sharp and a region where the color change is sharp on the image.

The extraction unit 19 has a color change degree comparison unit 21 which compares the color change degree obtained by the color change detection unit 13 with a reference value. The reference value can be arbitrarily set by inputting from a keyboard (not shown). The extraction unit 19 is provided with a region setting unit 23 that specifies the color change degree as a non-steep color change area when the color change degree comparison unit 21 determines that the color change degree is smaller than the reference value. The specification of the non-steep color change region means that the region where the color change is sharp is specified. When the reference value is increased,
The area where the color change is abrupt is small, and when the reference value is small, the area where the color change is abrupt is large.

Next, a pixel deleting unit 27 that receives the output of the region setting unit 23 is provided. The pixel deletion unit 27 deletes, from the input pixels, the pixels in the region where the color change is not abrupt at a constant rate. The ratio of deletion is
It is possible to set deletion: non-deletion = 1: 2 in the area. In addition, in other words, in the pixel deletion unit 27,
It may be said that some pixels in the non-steep area of color change are not used as data, and the steep portion is directly used as data.

The image data from which the pixels are deleted by the pixel deleting unit 27 is output as it is as compressed data, and the storage unit 29
Memorized in. When storing in this storage unit 29, ID
The assigning unit 25 may assign indexes (identifiers) to the non-deleted pixels and the deleted pixels. That is,
The first identifier is given to the pixel area that is the abrupt color change area. At the same time, the second identifier may be given to the pixel region in which the color change is not abrupt. As a method of assigning an identifier, for example, as shown in FIG. 3, “1” is previously assigned as an index to pixels in odd-numbered columns of pixels corresponding to each pixel of an image read by an image scanner, and an even number is assigned. “0” is given to the pixels in the column as an index in advance. Then, it is possible to exemplify a method of determining whether the pixel in the even-numbered column has a sharp color change with respect to the pixel in the odd-numbered column and inverting the index “0” to “1” if the color change is sharp.

According to this method, even after the color change judgment, since the indexes of the odd-numbered columns are all "1", it is not necessary to leave the index data, and it is sufficient to leave only the indexes of the even-numbered columns. Can save

The method of assigning the index is not limited to the above, and when the color change is abrupt, "1" and "1" are added to the successive pixels respectively, and in the pixel area where the color change is not abrupt. Alternatively, “0” may be added to the address of the deleted pixel and “1” may be added to the pixel that is not deleted.

Further, as shown in FIG. 4, the ID assigning section 25
An index may be added before pixel deletion by the pixel deletion unit 27. That is, in the area setting unit 23,
An identifier is given to a pixel region that is defined as a steep color change region and / or a non-steep color change region before pixel deletion. In this case, for example, in FIG. 3, the pixel to which “0” is added is deleted.

The compressed data is read from the storage unit 29 and restored by the decompression unit 31 as needed. Restoration unit 31
Has an ID reference unit 33 that determines whether or not an index is added to the area, or whether or not the added index is an index indicating a steep color conversion area. The restoring unit 31 leaves the portion determined to be the sharp color change portion by the ID reference unit 33 as it is, and copies the pixel data adjacent to the region where the pixel is deleted in the region which is the non-steep color change portion. It has a copying unit 35.

More specifically, at the time of image restoration, "1" is put in the odd columns in FIG. 3 and the stored index data is applied to the even columns. Looking at the indexes restored in this way, the indexes of "1" and "0" remain. The pixel having the index “1” is reproduced as it is, and the adjacent pixel is copied to the portion of the index “0”.

Further, the restoration unit 31 reproduces the image data restored by copying in the copying unit 35.
Have 7. The image data reproduced by the reproducing unit 37 is reproduced on the CRT display 39, for example.

An example of performing the image compression processing in the present embodiment having the above configuration will be described with reference to the flowchart of FIG. First, image data is read by an image scanner or TV camera (step 101). The obtained image data is decomposed into RGB signals by a color image analyzer and represented by 256 gradation density values (step 10).
2).

Next, from the RGB values of two adjacent pixels, the above-mentioned (R ² + G ² + B ² ) and (R ′ ² +) are calculated.
G ^'2 + B' ²⁾ for calculating a. The difference between the two is K = (R ² + G ² + B ² ) − (R ′ ² + G ′ ² +
B ^'2) is calculated (steps 103). Then, it is judged whether or not this K is larger than the reference value (step 104).

Here, the case where K is larger than the reference value is the case where the second pixel has a larger color change than the first pixel. In this case, both pixels are left as they are (step 105), and "1" and "1" are added as an index to the address table corresponding to both pixels (step 106).

When it is determined that K is smaller than the reference value, it is determined that the second pixel has less color change than the first pixel. In this case, the second pixel is deleted, leaving only the first pixel (step 107).
"1" is given as an index to the address table corresponding to the first pixel, and "0" is given as an index to the address table corresponding to the second pixel (step 108). The obtained compressed data is stored in the storage unit together with the index data (step 109).

The above processing is performed for all the fetched pixels (step 110), and the compression processing ends when the processing for the last pixel ends. At the time of restoration, as shown in the flowchart of FIG. 6, first, the stored compressed data and index data are stored in the storage unit 29.
Read out (step 121). With reference to this index data (step 121), for the part corresponding to the deleted pixel = index “0”, the adjacent pixel data is copied and reproduced (steps 123 and 125), and the part corresponding to the index “1”. Is reproduced as it is (steps 124 and 125).

FIG. 7 is a diagram showing an example in which the image is compressed and reproduced by the above method, and it can be seen that the reproduction degree in FIG. 7C at the time of reproduction is higher than that in FIG. 7A. On the other hand, FIG.
Shown below is a comparative example in the case where every second pixel is deleted, compressed, and reproduced. In this comparative example, the image of (C) at the time of reproduction becomes unclear as compared with (A).

The case according to the present invention and the case according to the comparative example are conceptually shown in FIGS. 9 and 10. <Embodiment 2> Next, a second embodiment of the present invention will be described.

In this embodiment, the color change degree output unit 17
The processing in is different. First, it is assumed that the RGB values recognized by the color image analyzer 15 are as follows.

First pixel: R ₁ , G ₁ , B ₁ Second pixel: R ₂ , G ₂ , B ₂ Third pixel: R ₃ , G ₃ , B _{3 In} this case, {(R ₁ + R ₃ ) / 2} -R ₂ = R _d {(G ₁ + G ₃ ) / 2} -G ₂ = G _d {(B ₁ + B ₃ ) / 2} -B ₂ = B _d Difference values R _d , G _d , and B _d between the second pixel and the intermediate point of the color change from the pixel to the third pixel are obtained. If this difference value is smaller than the preset reference value, the data of the second pixel is discarded. When the difference value is larger than the preset reference value, the data of the second pixel is left as it is.

Here, the values of R _d , G _d and B _d are respectively compared with the reference value, and if at least one of them is within the reference value (range of error), at least two of them are the reference values. Within, or at least three of which are within the reference value, or R _d + G _d + B
It is determined whether _d = K is within the reference value.

If it is within the reference value, the pixels in the area are deleted according to a fixed ratio. When it is larger than the reference value, the data in that area is left as it is. This process is the same as in the first embodiment.

FIG. 11 shows a flow chart in the case of this embodiment. Basically, it is the same as in the first embodiment.

[0051]

According to the present invention, by virtue of the above construction, it is possible to compress image data that can be reproduced naturally, which is not possible with the conventional compression method.

[Brief description of drawings]

FIG. 1 is a principle diagram of the present invention.

FIG. 2 is a block diagram showing a first embodiment of the present invention.

FIG. 3 is a diagram showing an example of assigning indexes.

FIG. 4 is a block diagram showing a modification of the first embodiment.

FIG. 5 is a flowchart showing an image compression process of the first embodiment.

FIG. 6 is a flowchart showing an image restoration process of the first embodiment.

FIG. 7 is a diagram showing an example of image compression / decompression according to the first embodiment.

FIG. 8 is a diagram showing an image compression / decompression example of a comparative example.

FIG. 9 is a conceptual diagram showing an example of image compression / decompression according to the first embodiment.

FIG. 10 is a conceptual diagram showing an image compression / decompression example of a comparative example.

FIG. 11 is a flowchart showing the image compression process of the second embodiment.

[Explanation of symbols]

101 ... Pixel reading step 103 ... Color change detecting step 104 ... Region extracting step 107 ... Pixel deleting step 108 ... ID assigning step 11 ... Input means 13 ... Color change detecting means 15 ... RGB Color image analyzer as value detecting unit 17 ... Color change degree output unit 19 ... Extracting means 25 ... ID assigning means 27 ... Deleting means 29 ... Storage means 31 ... Restoring means 33. ..ID reference section 35 ... Copying section

 ─────────────────────────────────────────────────── ─── Continued Front Page (72) Inventor Shigeki Furuta 1015 Kamiodanaka, Nakahara-ku, Kawasaki City, Kanagawa Prefecture, Fujitsu Limited (72) Inventor Hitoshi Matsumoto, 1015, Kamedotachu, Nakahara-ku, Kawasaki City, Kanagawa Prefecture, Fujitsu Limited

Claims (21)

[Claims] 1. A step (101) of reading an image as a plurality of pixels, a color change detecting step (103) of detecting a degree of color change between adjacent pixels, and a color change of a pixel in comparison with other areas. An image compression method comprising an extraction step (104) for extracting a region where the image is not sharp and a deletion step (107) for deleting some of the pixels of the region obtained in this extraction step. 2. The color change detecting step according to claim 1, wherein at least one of the R value, the G value, and the B value of the pixel is detected, and the detected R value and the G value are detected in adjacent pixels.
At least one of the B values is compared with each other, and the difference between the two is output as a color change degree, and the extraction step determines if the color change degree obtained in the color change detection step is smaller than a reference value. An image compression method characterized by specifying as a non-steep area of color change. 3. The method of claim 2, wherein the color change detecting step, as the degree of color change, and detects the value of R ² + G ² + B ² pixels, the same as R ² + G ² + B ² of the first pixel An image compression method characterized by comparing R ² + G ² + B ^{2 of} adjacent second pixels and outputting the difference as a color change degree. 4. The method of claim 2, wherein the color change detecting step, as the degree of color change, and detects the value of R ² + G ² pixels, the second adjacent and R ² + G ² of the first pixel and that Image compression method, characterized in that R ² + G ² of the pixel is compared and the difference is output as the degree of color change. 5. The method of claim 2, wherein the color change detecting step, as the degree of color change, and detects the value of R ² + B ² of the pixel, a second adjacent and R ² + B ² of the first pixel and that The image compression method, characterized in that R ² + B ² of the pixel is compared and the difference is output as a color change degree. 6. The method of claim 2, wherein the color change detecting step, as the degree of color change, and detects the value of G ² + B ² of the pixel, a second adjacent and G ² + B ² of the first pixel and that The image compression method characterized by comparing G ² + B ² of the pixels of the above and outputting the difference as the degree of color change. 7. The color change detecting step according to claim 2, wherein the RGB values of the first pixel are: R ₁ , G ₁ , and B ₁ The RGB values of the second pixel are: R ₂ , G ₂ , respectively. , B _{2 When} the RGB values of the third pixel are R ₃ , G ₃ , and B ₃ , respectively, the degree of color change is {(R ₁ + R ₃ ) / 2} −R ₂ = R _d {(G ₁ + G ₃ ) / 2} -G ₂ = G _d {(B ₁ + B ₃ ) / 2} -B ₂ = B _{d From} the calculation formula, the intermediate point of the color change from the first pixel to the third pixel, An image compression method characterized by obtaining and outputting difference values R _d , G _d , and B _d with respect to a second pixel. 8. The extracting step according to claim 7, wherein the values of R _d , G _d and B _d obtained in the color change detecting step are respectively compared with a reference value, and (1) at least one of them is a reference value. If (2) at least two of them are within the reference value, (3) If at least three of them are within the reference value, (4) or R _d + G _d + B
An image compression method, characterized in that it is judged that the color change is not abrupt in any case where _d = K is within a reference value. 9. The image compression method according to claim 1, further comprising an ID assigning step (108) for assigning an identifier to the non-steep color change region identified by the extracting step. 10. The extracting step according to claim 9, wherein when the color change degree obtained in the color change detecting step is larger than a reference value, the color change steep area is specified, and the ID assigning step includes: An image compression method characterized in that an identifier is given to each of the non-steep color change region and the steep color change region obtained in the extraction step. 11. An input means (11) for reading an image as a plurality of pixels, a color change detecting means (13) for detecting a degree of color change between adjacent pixels, and a color of a pixel as compared with other areas. Extraction means (1) for extracting a region where the change is not abrupt
An image processing device comprising 9) and a deleting means (27) for deleting some of the pixels in the area obtained by the extracting means. 12. The storage means (2) according to claim 11, which further stores pixels remaining after the deletion of the pixels by the deletion means.
An image processing apparatus comprising 9) and a restoration unit (31) for reading and restoring the pixel information stored in the storage unit. 13. The RGB value detection unit (15) according to claim 11, wherein the color change detection means detects at least one of the R value, the G value, and the B value of the pixel, and an adjacent pixel, A color change degree output unit (17) that compares at least one of the detected R value, G value, and B value and outputs the difference between the two as a color change degree; A color change degree obtained by the color change detecting means and a color change degree comparing section for comparing with an arbitrarily set reference value, and the color change degree comparing section judges that the color change degree is smaller than the reference value. An image processing apparatus characterized by specifying this as a non-steep area of color change when being processed. 14. The RGB value detection unit of the color change detection unit according to claim 11, detects the value of R ² + G ² + B ² of the pixel as the degree of color change, and the color change degree output unit is 1
R ² + G ² + B ² of the second pixel and R ^{2 of} the second pixel adjacent to it
An image processing apparatus characterized by comparing + G ² + B ² and outputting the difference as a color change degree. 15. The color change detection unit according to claim 11, wherein the R value of the pixel is determined as the degree of color change of the RGB value detection unit.
Detecting the value of ² + G ^2, the color change degree output unit, the difference as a color change degree by comparing the R ² + G ² of the second pixel adjacent the R ² + G ² of the first pixel and that An image processing device characterized by outputting. 16. The RGB value detection unit of the color change detection unit according to claim 11, detects the value of R ² + B ² of the pixel as the degree of color change, and the color change degree output unit sets the first image processing apparatus and outputs the difference as a color change degree by comparing the R ² + B ² of the second pixel adjacent the R ² + B ² to that of the pixel. 17. The RGB value detection unit of the color change detection unit according to claim 11, detects the value of G ² + B ² of the pixel as the degree of color change, and the color change degree output unit includes the first image processing apparatus and outputs the difference as a color change degree by comparing the G ² + B ² of the second pixel adjacent the G ² + B ² to that of the pixel. 18. The RGB value detecting section of the color change detecting means according to claim 11, wherein the RGB values of the consecutive first to third pixels detected as the degree of color change are the first pixel: R ₁ , G ₁ , B ₁ Second pixel: R ₂ , G ₂ , B ₂ Third pixel: R ₃ , G ₃ , B ₃ are detected, and the color change output section determines the degree of color change. , {(R ₁ + R ₃ ) / 2} -R ₂ = R _d {(G ₁ + G ₃ ) / 2} -G ₂ = G _d {(B ₁ + B ₃ ) / 2} -B ₂ = B _d An image processing apparatus characterized by calculating and outputting difference values R _d , G _d , and B _d between an intermediate point of color change from the first pixel to the third pixel and the second pixel from a calculation formula. 19. The image processing apparatus according to claim 11, further comprising ID assigning means (25) for assigning an identifier to the color change non-steep area specified by the extracting means. 20. The extracting device according to claim 19, wherein when the color change degree obtained by the color change detecting device is larger than a reference value, the extracting device specifies the steep color change region, and the ID assigning device, An image processing apparatus, wherein an identifier is assigned to each of the non-steep color change region and the steep color change region obtained by the extraction means. 21. The storage device (29) according to claim 19, further comprising: a storage unit (29) for storing pixels remaining after the deletion by the deletion unit, and a restoration unit (31) for reading and restoring the pixel information stored in the storage unit. And a pixel adjacent to a portion where the pixel is deleted based on the index referred to by the ID reference unit (33), which refers to the index assigned by the ID assigning unit. An image processing apparatus having a copying unit for copying data, and reproducing image data restored by copying.